Survival rates for oral cancer patients have remained unchanged in the past several decades largely because of diagnosis at late stage and local recurrence after treatment. In retrospective studies, patterns of loss of heterozygosity (LOH) have been associated with risk of progression of oral premalignant lesions to cancer and disease recurrence. The objective of the current proposal is to evaluate the prognostic value of specific LOH patterns in combination with clinical and histological features to predict outcome for oral lesions within a longitudinal study. We have used the unique medical infrastructure in British Columbia to establish one of the largest cohort studies of precancer and cancer patients in order to systematically follow changes in clinical, pathological and molecular parameters over time. The proposed grant renewal will extend the follow-up time of the established cohort and will accrue additional patients to 200 oral cancer patients and 200 patients with primary oral dysplasia. It will also fund the collection of tissue samples (biopsies, scrapes, brushings) and the analysis of allelic loss in these specimens. This extension will allow for better longitudinal modeling of risk patterns, including temporal changes of clinical, pathological and molecular markers, and will provide a multi-faceted risk model with clinical application that can be used to manage oral lesions and cancers.